EstoniaWe did ask if -ns is okay and assumed it is after the run got verified by Teo.
Edit: Oh yeah, but to be fair Teo didn't notice that the run used -ns (at least that's what he said in my chat when I asked about why he verified it).
I don't think it's worth messing up all other categories just because of p1 sorc. IGT never created fairness in p1 sorc which it was supposed to do. Beating Slimo -ns with IGT is not difficult, because once you get a Jail setup you already have a 3 minute head start. Had much more fun trying to beat it in -ns as well. I know 327 also enjoyed -ns.
IGT didn't make p1 sorc fair and just made all other categories unnecessarily more convoluted.
Not sure how to handle the -ns run. Slimo won the category with -ns. I'm personally fine with it just winning.
Is it a problem in any other category besides p1 normal sorc?
Edit: I can't really find any other category where it would matter as well besides p1 sorc.
Also didn't we already dethrone Slimo wr anyway when we converted to IGT. I guess we scam either way.
To me the Vice City style leaderboard still seems more fair and better understandable. I think it's a mistake that current runs do not have RTA listed in their data. It's going to be impossible to fix this issue if speedrun.com is going to support this hacky setup in the future. The longer it stays like this the more data will need to be fixed in the future.
Currently users don't even submit their RTA at all so how are we going to get that data. Run by run?
If RTA is required for the run to be submitted then that RTA data should be referenced in speedrun.com data anyway.
What was the reason?
There are lack of times in one of the leaderboards because they really don't exist. It would be a much more accurate leaderboard. For example https://www.speedrun.com/gtavc Also get a much faster way of sorting the runs by clicking time without loads or time with loads. Fixes the API problem as well.
Was there a reason not doing it this way? 
Speedrun.com was not built to support (at least not entirely) the hacky timing method solution we have going on here.
First issue is that RTA submissions get denied, but when you make a submission it's set as default. This also causes a problem I've noticed where verifiers don't notice and this happens: https://www.speedrun.com/d2lod/run/zxdr018y IGT run is submitted as RTA even though it's IGT and RTA submissions aren't allowed. Problems like this should be fixed early if we care about the data.
Second issue is that API doesn't list time without loads. If this was resolved we could nicely combine speedrun.com and diablo.run data. Otherwise we can't pull RTA/IGT separately since they're all the same in API.
For example here's my 48 minute sorc run https://www.speedrun.com/api/v1/runs/me683xqz and it has RTA set to 48 mins, but it's actually the IGT. API lists all IGT as "null".
API data should be cleaned up and always considered in future rule changes. We'll never be able to clean up the categories or history page if the data is broken inside.
I think it would be a really good change. Could actually link to px/p8 categories then. This change should overall resolve the px/p1/p8 debate.
I think we also don't need to split every main category into softcore/hardcore. Maybe only separate them when they become more popular and there's a demand for it.
It would be a good move towards less categories if we just combine sc/hc in px/p8 normal and hell. That change would already be like 28 less categories.
All these game comparisons don't make sense. Games like SM64 and Diablo 3 are exactly how I would like to see Diablo 2 categories sorted. Diablo 3 is literally sorted in "players x".
Even though SM64 doesn't show fastest category as default, it still shows fastest time in each category. In Diablo II that's not the case and that's the argument muffineater is making.
It would be a fair to compare to SM64 if the category setup looked like this for Diablo II: 
Sound definitely lags the game. It's a common fix to use -ns to improve game performance. Saw a lot of people running -ns in Median XL, Path of Diablo and Project D2 to make the game run more smooth. Mods have a lot more animations flying around in maps and the improvements with -ns are very noticable in these mods.
Second PC LiveSplit only possible with RTA. For D2 speedruns LiveSplit has to be on same PC as the game is running in order to track IGT.
There are plugins that support second PC IGT. For example https://github.com/livesplitclient/LiveSplitClientPlugin We should probably set this up if we stick with the IGT stuff.
You can use diablo.run overlay on your streaming PC. Add http://diablo.run/indrek/@#overlay as a browser source, but change username to yours.
460 width and 120 height, but you can use different sizes if you wish. The colors can be edited through OBS with custom CSS. For example here's the style I use:
tldr: Best tcp/ip load time possible is 8.3 seconds and it can be reached consistently every time by not using more than 11 squares of items.
Since our timer doesn't pause during TCP/IP loads I've been recently looking into how to go through it as fast as possible, because the 15 second loading times have been miserable.
Diablo 2 has a hardcoded 250 frames wait during TCP/IP loading (source PhrozenKeep). This has been fixed with mods like BaseMod that reduce the wait to 0 frames.
I don't think bypassing the 250 frames is possible, but we can eliminate other lag that happens during the load.
Menu runs at 30 fps which means 250 frames is 8.3 seconds. When you use TCP/IP with a character that has no items you can reach 8.3 seconds already super consistently. 8.3 seconds is the fastest possible load time due to hardcoded limitations.
Idea was to find a limit of items at which point it's longer than 8.3 seconds. It appears to be 11 squares. When more than 11 squares of items are used the load time goes to 10 seconds already. The more items you have the longer the load time is. It can go over 30 seconds when you fill the character up.
Doesn't matter if you fill the 11 squares in inventory/stash or belt but having a belt equipped does matter. If an item is 3 squares it doesn't seem to use up 3 squares. I think it might be related to the way Diablo stores items starting and ending square. More research is needed here.
Level, skills, stats, gold doesn't seem to impact the load time. Classic or various video settings didn't impact loading times.
8 seconds load time compared to 15 seconds gives you couple more chances each run. With 15 second load times I could maybe pull off 4-5 resets before it was too hopeless, but with the 8 second loads I felt like I could pull off 10 sometimes. Also it feels less frustrating when it goes by twice as fast.
Some TCP/IP research rambling video
My OBS background is blue, the game is black until I alt+F4. You can see it turn blue once Alt+F4 does the job and game closes.
There's a common black screen crash that happens. Would a run with this crash also get rejected?
Like this for example:
Bender was also just able to pull it off, I'll include the clip here for further research skipped wave 1 at around 50 seconds:
More Bender clips:
MacroBioBoi clips: wave 1 skip: 2 skips in a single run: Recovery skip: wave 2 skip: 2 waves in a row:
I was able to do it as well (wave 5 skip):
Original video that started this topic: skipping waves with blaze sorc clip (bnet)
Apparently this has also happened in Single Player (1.14) with an amazon
There must be a solid method out here to skip the baal waves. I've been trying to replicate it, but no success so far. Need more people to try out things and I'm sure we can figure it out together.
I also support adding any% -act5
